EVOKE Phase 3 trial results: Semaglutide for Alzheimer's disease: 3,808 patients with MCI or mild dementia. 104 weeks. Primary endpoint: no difference from placebo

Why do red light therapy studies contradict each other? Biphasic dose-response. Too little does nothing. Too much backfires. The sweet spot is narrow. Most consumer devices are not independently verified to hit it. The honest answer: we do not know if yours works. #medsky

Red light therapy has better evidence than most optimization interventions. The mitochondrial mechanism is plausible. Systematic reviews show modest benefits for muscle recovery and knee pain. Still not strongly recommended by major societies. But not dismissible either. #medsky

Peptide evidence: essentially zero adequate human trials. Red light therapy evidence: multiple systematic reviews, plausible mechanism, moderate effects for some conditions. This is what engaging seriously with the evidence looks like. Not all optimization is equal. #medsky

"Can I reverse my plaque?" Less important than you think. Plaque regression is a surrogate endpoint. A smaller unstable plaque is worse than a larger stable one. Statins regress plaque AND reduce mortality. Focus on outcomes, not imaging.

"I want the PCSK9 inhibitor, it is newer." You would not fund a startup with one Phase 2 trial when the competitor has 22 RCTs and mortality data. Why would the bar be lower for your coronary arteries?

When your kid brings home flu, Xofluza is one dose for contacts vs 10 days of Tamiflu. Sounds better. But households aren't single-exposure events. Ongoing exposure may need ongoing coverage.

In this flu season, we are reviewing cases of rhabdomyolysis with super high CK levels in the context of influenza A infection Here's a case report, but surprised there's not more in the literature: pmc.ncbi.nlm.nih.gov/articles/PMC...

Each day you delay flu treatment increases death risk by 14-40%. Both antivirals shorten illness by about one day if started within 24 hours. After 48 hours, Xofluza has no data at all.

If you care about healthspan, you have to care about blood pressure. Uncontrolled BP claims 10 million lives yearly. Intensive control cuts cardiovascular events by 30%. Taking a peptide feels sophisticated. Checking your BP feels ordinary. The evidence base isn't close.

The choice that matters isn't Tamiflu vs Xofluza. It's today vs tomorrow. Same-day flu treatment was associated with 64% lower mortality. Drug selection? Similar outcomes.

Beyond the headline: cystatin C discordance with creatinine eGFR is about more than just kidney function; it predicts outcomes, may suggest systemic problems, and the directionality is critical (GFRcr > eGFRcys is a good sign). buff.ly/CscgFAJ #medsky

Software can be patched. Markets recover from failed experiments. The body's irreplaceable systems cannot. "Move fast and break things" doesn't work here.

Patients overestimate statin side effects by an order of magnitude. True attributable muscle symptoms: less than 1%. Severe myopathy: less than 0.1%. The perception gap drives people toward less-proven alternatives.

How smart people fool themselves about health interventions: Subtle adjustments to probability estimates. Shifts in confidence levels. Quiet downgrades of source reliability for inconvenient findings. Small biases that accumulate into confident but unfounded conclusions.

The same traits that drive innovation—systematic thinking, high openness, confidence in analysis—become vulnerabilities when applied to human biology. The mindset that builds companies requires different guardrails when applied to health.

Statins reduce all-cause mortality in high-risk primary prevention (RR 0.92). PCSK9 inhibitors do not demonstrate that benefit in this population. The boring generic pill has the outcome data the fancy injectable lacks.

Dunning himself clarified his research: "You can be incredibly intelligent in one area and completely not have expertise in another." Tech founders approaching human biology with the confidence they bring to software development is a textbook case.

"Unintelligent people are more easily misled by others. Intelligent people are more easily misled by themselves." They construct more sophisticated rationalizations. Find creative ways to dismiss contrary evidence.

Aging muscle isn’t just smaller. It's different. Soendenbroe et al. 2024 (J Cachexia Sarcopenia Muscle): heavy resistance training reverses irregular myofiber shape, a marker of aging. Net: strength work improves muscle quality, not just size. For healthy aging, resistance training is non‑negotiable

I’m quoted in today’s @nytimes.com on the "Chinese peptides" trend in SF. She asked about the "founder mindset" applied to biology. My take: "The entrepreneurial parallel isn't funding a scrappy start-up. It's wiring money to an unregistered offshore entity based on a pitch deck." buff.ly/n5nEob

Creatinine-based eGFR estimation is the default, cheap, fast, and often wrong. Cystatin C is more accurate, must be ordered separately, and in many settings takes days to return. Why? #medsky

The simple pleasure of chewing on a stick in the morning.

I know it's been an eventful year but let's not forget that AI served up this monstrosity of a figure in a paper that was published and then retracted by one of the frontiers journals this year. The rat with giant balls lives in AI infamy for all time scienceintegritydigest.com/2024/02/15/t...

T-SPOT.TB performs much better in many immunocompromised populations IMO, but historically was logistically more difficult from CA due to time-sensitive FedEx of samples to Memphis which failed a lot with winter storms. Less an issue now that Quest bought T-SPOT.

Pet rocks are going to trend

Mabel is not a bad bunny (Trending topic)

Mabel and Merlin exploring their new hidey house.

My BSKY policy is to block people who seem to be trolls on sight.

Reading the lawsuit - just wow. Good on her for bringing this forward

Late night hop & flop action.

Friday's looking good to Mabel

Saturday morning in bunland.

CDC exercises caution but endorses doxycycline and tetracycline as viable alternatives, with a preference for doxycycline, given its better compliance rates and reduced incidence of gastrointestinal adverse effects. #idsky #medsky

Compared Bluesky vs. X for the UHC and tsunami news this morning. Bsky easily for the win on both.

Sunday Bunday #bunnies

Mabel says "TGIF" #bunnies

Updated with the ONE THING that Microsoft Teams does better than Zoom. Grateful for that! @ademeter.bsky.social @rbganatra.bsky.social

"Drink more water!" It's the easiest health advice around. Tired? Headache? Trying to be like Tom Brady? Just drink more water. But as a nephrologist, I get asked all the time: How much is the right amount? And the truth is... we don't really know. 🧵

Morning #bunnies complement my coffee nicely.

This is your reminder to get a library card. They have so many services and basically all of them are free.

"Disrupt" is code for "Destroy reliable established services so that money previously widely distributed now funnels into a single source." Just like "AI" is code for "Mine the work humans did in the past so we don't have to pay them for their work in the future." That's why VCs like both words.

Merlin's got the "binkies" on a Sunday morning.. #bunnies

Does this make #idsky digital benzathine penicillin G?

Nerd humor 🤓 #statisticshumor #cardiosky #medsky

I think when I break the news to my 50-something patients that pneumococcal vaccine is now recommended for them, they just hear “the CDC just declared that I’m old.” #medsky

Brunch is always better with happy bunnies #ASMR

Bunnies gonna bun.

🚨 Antibiotic Treatment for 7 versus 14 Days in Patients with Bloodstream Infections BALANCE trial publication is here ! #IDsky #Medsky @idjclub.bsky.social www.nejm.org/doi/full/10....